Summary Peace Air Ltd. Flight PE905, a British Aerospace Jetstream 3112 (registration C-FBIP, serial number820), was conducting an instrument approach to Runway29 at Fort St.John, British Columbia, on a scheduled instrument flight rules flight from Grande Prairie, Alberta. At 1133 mountain standard time, the aircraft touched down 320feet short of the runway, striking approach and runway threshold lights. The right main and nose landing gear collapsed and the aircraft came to rest on the right side of the runway, 380feet from the threshold. There were no injuries to the 2pilots and 10passengers. At the time of the occurrence, runway visual range was fluctuating between 1800and 2800feet in snow and blowing snow, with winds gusting to 40knots. Ce rapport est galement disponible en franais. Other Factual Information History of the Flight The flight departed Grande Prairie at 1040 mountain standard time1 on the first leg of a scheduled run to Fort St.John (CYXJ), Fort Nelson, British Columbia, and Grande Prairie. It was operated under Part704 of the Canadian Aviation Regulations (CARs). The captain was the pilot flying and was occupying the left seat. Air traffic control (ATC) cleared the flight to maintain 12000feet above sea level (asl), and to hold on the TAYLOR non directional beacon (NDB), which forms the final approach fix for the instrument landing system (ILS) approach to Runway29 at CYXJ (see AppendixA). At 1109, Flight PE905 was cleared for an ILSapproach to Runway29. The first approach was discontinued due to the aircraft being too high on the final approach leg, and clearance was given to return to the TAYLORNDB at 6000feet. At 1123, ATC cleared the flight for another ILSapproach to Runway29. The second approach was conducted as a full procedure with the outbound leg extended to ensure that the aircraft was positioned to follow the correct vertical approach profile. The final approach course was flown with a flap setting of 20 and at the company standard operating procedure (SOP) recommended airspeed of 130knots. At approximately 300feet above ground level (agl), the first officer informed the captain that he had the ground in sight. The approach lights were visual shortly thereafter. The captain discontinued his instrument scan and confirmed the appearance of the approach lights. The captain made the decision to land, and called for the full flap setting of 35. The first officer diverted his attention to setting flaps, and to the Vref2 reference cards clipped to the instrument panel. When he looked up, the aircraft was almost on the ground, but short of the runway. There was insufficient time to warn the captain. The aircraft initially touched down in a nearly wings-level attitude, 320feet short of the threshold in about 16 inches of packed snow. The left main gear contacted the surface first, followed by the right main gear, and then the nose wheel. The aircraft then struck the last set of approach lights, bounced slightly, and touched down again 180feet short of the threshold. After sliding through the threshold lights, the aircraft came to rest on the right edge of the runway, 380feet beyond the threshold. The right main gear had broken off, and the nose gear had collapsed rearward. Both propellers were damaged by ground contact. The aircraft was equipped with a belly-mounted cargo pod, which supported the fuselage during impact. After the aircraft came to a stop, the captain informed emergency services of the crash in a cellular telephone call to911, and the Aircraft Rescue and Fire Fighting service was activated. A snow-removal crew was first on the scene soon after the aircraft came to rest; emergency response vehicles arrived within 15minutes. Due to the poor visibility and high wind chill, the occupants remained in the aircraft with the door closed until they could be transferred to the vehicles. All of the passengers were wearing lap belts during the approach and landing, and were retained in their seats. Most of the passenger baggage was loaded in internal and external cargo compartments, and few carry-on items were in the cabin. There were no passenger injuries due to movement of unsecured items during the ground impact and subsequent deceleration. The emergency locator beacon activated automatically on landing, and it was turned off manually. Weather While Flight PE905 was conducting its approaches, the weather at CYXJ was deteriorating rapidly in blizzard conditions associated with the passage of a warm front aloft. En route to CYXJ, the crew obtained the weather on the automatic terminal information system (ATIS). Message "N," which was based upon a special weather observation at 1020, stated the following: wind 360 true (T) at 10knots, visibility 1to 3statute miles (sm), vertical visibility 2300feet, temperature -8C, dew point -9C, altimeter 29.65, runway 80percent bare and dry. The 1100 regularly scheduled observation was as follows: wind 350T at 15knots gusting to 25knots, visibility variable from sm to 1sm in light snow and drifting snow, Runway29 runway visual range (RVR) was 3500feet variable 5000feet and trending downward, vertical visibility was 1100feet with remarks of snow obscuring 8oktas3 of the sky. A special observation at 1125, 8minutes before the accident, showed the wind increasing further to 340T at 30knots gusting to 40knots, visibility was variable 0to sm in snow and heavy blowing snow, Runway29 RVR was 1800feet variable 2800feet and stable, vertical visibility was 400feet, and remarks of snow 8oktas. Two revised ATIS messages, "O" and "P," were issued based on the 1100and 1125observations respectively. The crew did not tune the ATISfrequency to receive these messages. Information regarding the 1125observed visibility of 0to sm was not passed to the crew. When the aircraft was on final approach, nine minutes before the landing, the Flight Service Station (FSS) informed FlightPE905 that the wind was 310at 30knots gusting to 40knots, the sky was obscured, RVR was 2800feet, and runway lights were at the full-intensity setting of strength five. The RVR of 2800feet was greater than the CAR approach ban limit of 2600feet for commercial aeroplanes conducting precision approaches. The weather observation point at the FSS and the Runway29 RVR transmissometer are about 0.8nautical miles (nm) apart. The final stage of the approach to Runway29 was across a snow covered field with little visual contrast. The crew of a scheduled air carrier flight, which departed CYXJ 13minutes before Flight PE905landed, reported marginal weather conditions. This report was made on the Area Control Centre (ACC) frequency, and since the crew of Flight PE905 was monitoring the mandatory frequency, they did not hear that information. The report was not relayed to FlightPE905. There were no indications of ice accumulation on the aircraft before or after the accident. Flight Crew The pilots were properly licensed in accordance with existing regulations. Their flight and duty times met regulatory requirements, and they were considered to be well rested. The captain held an airline transport licence and was employed by the company since May2006. His total flying time was 13000hours, with 300hours on type and about 450hours in instrument flight rules (IFR) operations. The majority of his flying hours were under visual flight rules (VFR), with extensive seaplane experience. The first officer held a commercial pilot licence, and was employed by the company since September2006. His total flying time was 275hours, with 20hours on type. This was the first officer's first operational instrument approach in instrument meteorological conditions (IMC) on the aircraft type. Company Procedures The approach briefing by the captain did not include information pertaining to flap selection. Changes in the flap setting result in changes to Vref. The company's SOPs for a precision instrument approach on the Jetstream3112 indicated that flaps should be extended to 35 (full flap) no later than crossing the final approach fix. The SOPs also stated that the aircraft should be stabilized in airspeed, landing configuration and descent rate by 500feetagl. Selection of full flap at an airspeed of 130knots results in the aircraft pitching up. In order to avoid a climb, the control yoke must be moved forward, to pitch the nose down.4 A bound set of approach charts was available for the crew's use. A photocopied set of charts was positioned in front of the first officer, who passed relevant information to the captain as the approaches progressed. The captain did not refer directly to approach charts during the approaches. The first officer read the value for the decision height5 (DH) as 2400feetasl, rather than the published 2454feet. The captain did not confirm the accuracy of the information. A cold temperature correction factor6 was not applied to altitudes during the approach. The charted cold temperature correction for -8C at 200feet agl was an additional 20 feet to the DH. The pilot monitored approach (PMA) is a common industry procedure used on instrument approaches in low weather conditions; the first officer normally flies the approach while the captain monitors the instruments. Approaching minimums, the captain begins to look outside for the appropriate visual cues. At DH, the captain will take control and land if he has the appropriate visual cues, or he will have the first officer continue on instruments until the appearance of more visual cues and then take control and complete the visual landing. When control is transferred, the first officer continues to monitor the flight instruments until touchdown. If visual references are lost at any time, the captain would command a missed approach, and the first officer would fly the missed approach procedure. The PMA affords a continuity of instrument monitoring during the critical phase of an instrument approach close to the ground in low ceiling and visibility. It enables the first officer to remain on instruments until touchdown to alert the captain of any small pitch change that could give rise to a significant change in the rate of descent or airspeed. The company's SOPs did not provide for PMAs, but indicated that the pilot who flies the approach would also carry out the landing. The pilot not flying would monitor the performance of the approach; call airspeeds in relation to Vref; call altitudes, and manage aircraft systems according to checklists; set instruments and avionics; and handle radio communications. Fort St. John Airport Information Runway 29 at CYXJ is 6900feet long and 150feet wide. It is equipped with a high-intensity approach light system with sequenced, flashing lead-in lights, and runway alignment indicator lights. The approach light system is 2400feet long, and terminated by a bar of green threshold lights. Two hundred feet separate each system component. The runway is served by an ILS with a glide slope angle of 3.0, which places an aircraft at a height of 50feet when crossing the threshold. DH for the ILS approach is 2454feet asl, or 200feet agl. The published landing visibility7 is sm or RVR 2600feet (see AppendixA). Required visual reference is defined as the section of the approach area of the runway or visual aids including approach lights that, when viewed by the pilot of the aircraft, enables the pilot to make an assessment of the aircraft position and the rate of change of position relative to the nominal flight path. Aircraft Information The aircraft was reported to operate normally throughout the flight, and records indicated that it was certified and maintained in accordance with Transport Canada regulations. It was not equipped with an autopilot. A serviceable flight director was installed, but was not used by the crew on this flight. Aircraft weight and balance was calculated to be within limits published in the aircraft weight and balance manual. A flight data recorder was not installed, nor was one required by Transport Canada regulations. Cockpit voice recorder (CVR) data from the last 30minutes of the flight were retrieved by the TSB Engineering Laboratory. During the accident flight, the CVR did not record the first officer microphone or hot microphone channels. However, first officer voice information was recorded on the captain's hot microphone intercom channel. A bench test of the CVR determined that the fault in the first officer microphone channel likely originated in the airframe audio system. An annual intelligibility test, as required by the CARs, was conducted on the CVR by an avionics shop in May 2006. At that time, it was noted that the first officer's press-to-talk system was satisfactory and the hot boom microphone was unsatisfactory. Overall, the test was evaluated as "marginal." The CVR was returned to service with no further work done on the system. It is the operator's responsibility to interpret the test results, and to determine whether the CVR meets the regulatory standards before returning the CVR to service. Section 625.33 of the CARs states that CVRs shall continuously record the following: the audio signals received from each microphone being used by a flight crew member; and voice communications of flight crew members using the aircraft's interphone system. the audio signals received from each microphone being used by a flight crew member; and voice communications of flight crew members using the aircraft's interphone system.